Building and managing cloud services at scale is new to most of us; we’re either building our first tranche of cloud-native applications or have started to migrate existing applications from on-premises to the cloud. In many cases we haven’t had time to prototype and pilot. We’re being driven to rapidly take businesses digital by a global pandemic and a sudden shift to remote working.
The question then is: What are the best practices for working with hyperscale clouds such as Azure? What worked well in on-premises data centers may not be a good fit for virtual infrastructures or container-based microservice orchestration using Azure. Then there’s the question of how to size and scale for the cloud, where compute resources may well be unlimited but they’re not free, and we have to account for every byte of storage and every second of CPU time.
Introducing the Azure Well-Architected Framework
Microsoft’s developer relations and professional services groups have been working to help newcomers to the cloud (as well as anyone who wants to improve existing applications). It recently published its guidelines as the Azure Well-Architected Framework, codifying best practices and tying them into the business needs of organizations beginning a migration to the cloud.
The framework covers five key areas: managing costs, implementing an operations model, ensuring efficient performance, application reliability, and running securely. All five are part of running any application infrastructure anywhere, but each has uniquely cloud-centric requirements. Running a cloud service needs a different mix of skills, especially as it leads to having separate cloud operations, virtual infrastructure operations, and application operations teams rather than a single data center operations group.
Running an assessment
At the heart of the Azure Well-Architected Framework is a self-guided review to help you assess where you are in your cloud migration and how you see your priorities. Be honest in your assessment. Otherwise you’ll get a false picture of your current state which could lead to problems in the future.
Each section of the review is a brief multiple-choice quiz. Each selection roll-over details a quick tip that can help with fulfilling any requirements. It’s a useful approach, as working your way through the questionnaire may get you thinking about steps you haven’t yet taken and wish to investigate further. It’s clear that the aim of the assessment is as much to get you thinking in a cloud-centric way as it is to collect data to help you design better applications.
Once you’ve completed the review you’re presented with a score out of 100 that indicates how close you are to the ideal architectural approach. Ratings for individual areas are shown, giving you an idea of what should be prioritized. Recommended actions are provided as links to relevant documentation, helping you consider what can be done with your application now and in future. Most of the documentation is in the Azure Architecture Center, which is Microsoft’s hub for Azure design patterns and best practices.
This article was originally published by InfoWorld. Algoworks does not take any credit and is not responsible for the information shared in the article.
